In short: Schema.org structured data tells search engines exactly what a page contains: an article, a product, an organization or a FAQ. Markup unlocks rich snippets and helps AI search engines quote your content accurately. JSON-LD is the recommended format. Check structured data online with the schema.org validator, Google Rich Results Test, Yandex.Webmaster or our schema-checker.
What structured data is and why it matters
Structured data (schema markup) is a machine-readable description of a page based on the Schema.org vocabulary. Regular HTML tells the browser how to render text, but it never explains to a crawler that "4.8" is a product rating and "$129" is a price. Schema markup removes the guesswork: you explicitly declare the entity type and its properties, and the search engine stops interpreting your page by indirect signals and heuristics.
The practical value comes down to three effects. First, rich results: star ratings, price and availability, FAQ blocks, breadcrumbs instead of a raw gray URL. A rich snippet occupies more space on the results page and reliably lifts CTR even when positions stay the same. Second, eligibility for special SERP features: article carousels, product grids, knowledge panels. The third effect has become critical over the last few years: AI search engines and LLM assistants quote pages with structured data far more willingly, because facts — the author, the publish date, the organization name, product specs — are trivial to extract from a JSON-LD block and expensive to guess from raw HTML.
Structured data does not boost rankings directly, but it changes how your page looks in search. A snippet with a rating and a price collects more clicks than the plain blue link next to it — and click-through behavior does feed back into ranking.
Formats: JSON-LD vs microdata vs RDFa
The vocabulary is the same — Schema.org — but there are three syntaxes for embedding it. The difference matters mostly for maintainability: two formats are woven into your HTML tags and live inside the layout, while the third sits in a separate block and does not depend on the markup at all.
| Criterion | JSON-LD | Microdata | RDFa |
|---|---|---|---|
| Placement | Separate block in head or body | Attributes inside HTML tags | Attributes inside HTML tags |
| Coupling with layout | Independent of the markup | Breaks when templates change | Breaks when templates change |
| CMS generation | Simple: one script template | Hard: edits across all templates | Hard |
| Search engine support | Google, Yandex, Bing | Google, Yandex, Bing | Partial |
| Recommendation | Yes, the primary format | Legacy templates only | Not recommended |
Google explicitly recommends JSON-LD in its structured data documentation: the script does not interfere with your layout, it is easy to generate from a template or CMS, and easy to test in isolation. Microdata is only justified when your CMS already ships templates with itemprop attributes and rewriting them would cost more than it returns. RDFa survives mostly in legacy projects.
Core Schema.org types for a website
Article and its subtypes BlogPosting and NewsArticle describe articles and blog posts. Key properties: headline, description, image, author, publisher, datePublished and dateModified. Google uses them for article carousels and news blocks, while AI search engines use them to attribute the author and the date correctly when quoting your text.
Product is a product card: name, image, description, plus an offers object with price, currency and availability. Combined with AggregateRating it produces the classic snippet with stars and a price. For an online store this is the highest-ROI markup type: extra CTR on product pages converts directly into orders.
FAQPage is a list of questions and answers. Each question is a Question entity with an acceptedAnswer property. Important: questions and answers must be visible on the page — markup hidden from users is treated as spam. FAQ markup is heavily consumed by AI assistants: ready-made question-and-answer pairs are the easiest content to quote.
Organization (and LocalBusiness for physical locations) holds company data: name, logo, address, phone, links to company profiles elsewhere via sameAs. It helps search engines build a knowledge panel and connect your site to the company profiles on other platforms. Place it once — usually on the home page or the contacts page.
BreadcrumbList renders breadcrumbs in the snippet, replacing a gray URL with a readable path like Catalog → Windows → PVC. One of the simplest types to implement and one of the most visible in search — especially on mobile, where long addresses get truncated.
HowTo is a step-by-step instruction: steps, time, tools and materials. It fits manuals and tutorials. Google has scaled back HowTo rich results on the SERP, but the markup still helps structure content for AI search engines that retell instructions step by step.
The full vocabulary contains hundreds of types: events, job postings, recipes, videos. The up-to-date list lives at schema.org.
How to implement structured data: JSON-LD examples
To add structured data, place a script block with the application/ld+json type in the head or body — the position does not matter, crawlers read both. The cleanest scheme is one entity per block; related entities can be combined via @graph. A minimal example for an article:
<script type="application/ld+json">
{
"@context": "https://schema.org",
"@type": "Article",
"headline": "Schema.org Structured Data: How to Check and Implement It",
"description": "A guide to implementing and validating structured data.",
"image": "https://example.com/assets/cover.webp",
"author": {
"@type": "Person",
"name": "John Smith",
"url": "https://example.com/authors/john"
},
"publisher": {
"@type": "Organization",
"name": "Enterno",
"logo": {
"@type": "ImageObject",
"url": "https://example.com/logo.png"
}
},
"datePublished": "2026-07-21",
"dateModified": "2026-07-21"
}
</script>
Field notes: keep headline under 110 characters, use ISO 8601 dates, and make image an absolute URL — relative paths are a frequent silent failure. The publisher logo is required for article rich results in Google. The second example is a FAQ block:
<script type="application/ld+json">
{
"@context": "https://schema.org",
"@type": "FAQPage",
"mainEntity": [
{
"@type": "Question",
"name": "Does structured data affect rankings?",
"acceptedAnswer": {
"@type": "Answer",
"text": "Not directly, but a rich snippet raises CTR."
}
},
{
"@type": "Question",
"name": "Which markup format should I choose?",
"acceptedAnswer": {
"@type": "Answer",
"text": "JSON-LD: it is the format Google recommends."
}
}
]
}
</script>
For FAQPage the golden rule is parity: every question and answer in the markup must exist on the page as visible text, and the wording should match. Markup that promises content the user cannot see violates search engine guidelines.
How to check structured data: four validators
Validation takes five minutes and saves weeks of silent waiting: a crawler that hits broken markup simply ignores it — no warnings, no errors in your webmaster console. Four tools cover all the practical cases.
Schema.org validator (validator.schema.org) is the official vocabulary tool for checking structured data online by URL or by code fragment. It catches syntax errors and type mismatches: typos in property names, unclosed brackets, wrong nesting. It does not know the requirements of specific search engines — it is a grammar check, not a rich result eligibility check.
Google Rich Results Test shows which rich results the page is eligible for and flags the required properties without which the enhanced snippet will not render. It accepts both a URL and a raw code fragment — convenient for testing before deployment.
Yandex.Webmaster offers a structured data checker under its Tools section. It parses JSON-LD, microdata and RDFa and applies Yandex-specific requirements — relevant whenever part of your audience searches in Russian.
Our validator — check your structured data in schema-checker: paste a URL, and the tool finds every structured data block, detects the types and lists the errors. Markup is only one part of technical SEO: for the full picture across indexing, speed and meta tags, run a comprehensive SEO audit.
Do not validate a single URL — validate one page per template: the home page, a category, a product card, an article. A template error replicates across the whole site, and thousands of pages lose their rich snippet at once.
Common structured data errors
- Marking up data that is not on the page: a rating with no reviews, an FAQ with no visible questions. A direct guideline violation that can trigger a manual action.
- Missing required properties: Product without offers, Article without datePublished. The validator shows a warning and the rich result never appears.
- Relative URLs in image and logo — crawlers require absolute addresses with a protocol.
- Broken JSON: a trailing comma or an unclosed quote. One character kills the whole block.
- Markup drifting away from the content: the JSON-LD says $99 while the page says $149. Search engines stop trusting the domain data.
- Duplicates: two Organization blocks with different names on the same page — a common side effect of a theme plus an SEO plugin.
If your rich snippets disappeared or pages dropped out of the index, start with basic diagnostics: how to check a website for errors and why a site is missing from search.
FAQ
Does structured data affect rankings?
Not directly — it is not a ranking factor. Indirectly — yes: a rich snippet raises CTR, and better click-through improves behavioral signals. Markup also helps the engine understand the page precisely and match it to more relevant queries.
Can I put several types on one page?
Yes, that is standard practice: an article can carry Article, BreadcrumbList and FAQPage at the same time. Use a separate JSON-LD block per type or combine them via @graph. Just never mark up the same entity twice in different formats.
How soon will rich snippets appear?
After the page is recrawled — usually a few days to a couple of weeks. There is no guarantee though: a rich result is a privilege, not a right; the engine decides whether to show the enhancement even for perfectly valid markup.
Is a CMS plugin enough?
For standard pages — yes: WordPress SEO plugins and similar CMS modules generate correct JSON-LD for articles and products. But verify the output anyway: plugins routinely duplicate blocks or emit empty fields.
Implementation checklist
- Pick types that match the content: Article for the blog, Product for product cards, Organization for the home page.
- Use JSON-LD, not microdata.
- Fill in every required property of the type plus the ones recommended for the rich result.
- Make sure the markup matches the visible page content exactly.
- Use absolute URLs for images and logos.
- Run one page per template through schema-checker and Google Rich Results Test.
- Add markup validation to your release routine: a template change can silently break it.